Bayesian Selection Of Grammar Productions For The Language Of Thought
Sergio Romano, Alejo Salles, Marie Amalric, Stanislas Dehaene, Mariano Sigman, Santiago Daniel Figueira
This paper demonstrates how to select grammar productions for a Language of Thought (LoT) using Bayesian inference, testing it in a geometric LoT model and finding an empirical relationship between sequence probability and complexity.
PLoS ONE · Jul 10, 2018Peer-reviewedNeuroscience